Can neighborhood design support walking? Cross-sectional and prospective findings from Japan

Objective: We examined (1) cross-sectional and prospective associations of perceived and objectively-measured neighborhood attributes with purpose-specific walking; (2) the differences between cross-sectional and prospective associations in the sample of Japanese middle-to-older-aged adults. Methods: We conducted a prospective cohort study to collect data from 544 adults aged 40 to 69 years living in two cities in Japan in 2011 and again in 2013. Generalized linear modelling was employed to examine associations of perceived and GIS-measured built environment attributes (population density, access to destinations, access to public transportation, sidewalk, and street connectivity) with four types of self-reported purpose-specific walking, namely walking for commuting (to and from work), walking during work, walking for errands (shopping, to the bank or post office), and walking for exercise. Results: After adjusting for potential confounders, GIS-measured higher population densityand better street connectivity were associated with increased time spent walking for commuting and exercise. Furthermore, GIS-measured better access to public transportation was related to increased time spent walking for commuting, and perceived better access to destinations was also associated with increased time spent walking for commuting and daily errands. Unexpectedly, GIS-measured better access to destinations, and perceived sidewalk presence were related to decreased time spent in work-related walking. Conclusion: Living in high-density neighborhoods with well-connected streets and convenient public transportation systems, and having a higher awareness of destinations are supportive of the long-term engagement in walking for various purposes. Further studies using a prospective design with longer follow-up period to confirm these results are warranted.